The purpose of the proposed research is the testing of different key aspects of the functional model of nonverbal exchange (Patterson, 1982). The basic focus is the predicted contrast between the intimacy and social control functions of nonverbal involvement. These functions will be examined in interactions in which expectancies about the subject's goals in an interaction will be varied. Specifically, it is predicted that when the subject expects a casual, informal interaction, the subject's nonverbal involvement with that partner will be determined more by the spontaneous affect (intimacy) toward the partner. In contrast, when a subject expects to create a favorable impression on the partner, the subject's nonverbal involvement with that partner will be determined more by the purposeful goal (social control) of creating a good impression. In the case of the intimacy function, there should be consistency between interpersonal affect and nonverbal involvement. That is, the more positively the subject views the partner, the greater should be the subject's level of nonverbal involvement. In contrast, in the case of the social control function, there will be relative independence between interpersonal affect and nonverbal involvement. That is, the subject's level of nonverbal involvement will be determined more by achieving the desired reaction from the partner than by the subject's affect toward the partner. The proposed research will examine not only the consequences of different expectancies on nonverbal involvement, but also the cognitive and physiological reactions that are presumed to mediate those expectancies. Such a multivariate approach should contribute to a better understanding of nonverbal exchange and the factors that mediate that exchange. Because nonverbal behavior is so important for understanding a interaction processes, this research is relevant for a wide range of basic and applied issues.